Ship welding quality management and control system and method

Abstract

The invention discloses a ship welding quality management and control system and method, and aims to solve the problems of data splitting, risk lag and adjustment blindness in traditional management and control. The system comprises a welding data acquisition module (real-time acquisition of welding electrical parameters and environmental parameters), a welder identity authentication module (RFID verification unlocking equipment), a pre-control module (core including a multi-source feature alignment unit, a risk level mapping unit and a dynamic instruction generation unit), a quality result filing module (MES system structured storage) and a management and control terminal interaction module (touch screen display of risks and instructions). The method comprises the steps of welder authentication, multi-dimensional acquisition, pre-control processing, real-time feedback and quality filing.

Description

Technical Field

[0001] This invention relates to the field of shipbuilding quality control, specifically to a ship welding quality control system and method. Background Technology

[0002] Ship welding is a core process in shipbuilding, and its quality directly affects the structural strength of the hull and navigational safety. Traditional ship welding quality control faces the following key problems:

[0003] Data fragmentation: Welding current, ambient temperature and humidity, welder skills, etc. are collected separately without establishing a correlation, making it impossible to determine the "quality risk under the combined effect of multiple factors" (such as "high humidity + low current" easily leading to incomplete fusion).

[0004] Risk lag: Relying on UT / RT flaw detection after welding is completed to detect defects means that by the time problems are discovered, material waste and project delays have already occurred, and real-time intervention is impossible;

[0005] Blind adjustments: Even if real-time parameters are found to be abnormal, adjustments are made solely based on the welder's experience (such as "adjusting the current if it is too high") without taking into account environmental factors, skill level, and other factors to provide a precise adjustment range, which can easily lead to new defects.

[0006] Therefore, there is an urgent need for a control solution that can integrate multi-dimensional data, pre-control quality risks in real time, and generate precise adjustment instructions. Summary of the Invention

[0007] The purpose of this invention is to provide a ship welding quality control system and method to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a ship welding quality control system, comprising a welding data acquisition module, a welder identity authentication module, a pre-control module, a quality result archiving module, and a control terminal interaction module; the welding data acquisition module, through current, voltage, molten pool temperature sensors and workstation environment sensors integrated into the welding equipment, collects welding electrical parameters (80-500A current, 15-40V voltage, etc.), molten pool state (800-1500℃), and environmental parameters (temperature, humidity, wind speed) in real time, and outputs standardized raw data at a sampling frequency of ≥50Hz;

[0009] The welder identity authentication module reads the welder's identity information (certificate level, preferred welding position, historical pass rate) through an RFID chip to verify whether the welder meets the skill requirements of the current workstation (e.g., vertical welding requires a Level III welder). If qualified, the welding equipment is unlocked. At the same time, the welder's skill parameters (especially the historical pass rate) are synchronized to the pre-control module.

[0010] The pre-control module receives real-time process data from the welding data acquisition module and skill data from the welder identification module. First, a multi-source feature real-time alignment unit performs data time calibration, dimension unification, and invalid data removal. Then, a welding quality risk level mapping unit, combined with a multi-factor weighting model, calculates risk coefficients, classifies risk levels (green / yellow / red), and locates risk sources. Finally, a dynamic adjustment instruction generation unit outputs precise adjustment suggestions based on the risk sources. The output risk level and adjustment instructions are simultaneously pushed to the control terminal interaction module, while the process data is transmitted to the quality result archiving module. The pre-control module includes a multi-source feature real-time alignment unit, a welding quality risk level mapping unit, and a dynamic adjustment instruction generation unit. It can integrate welding electrical parameters, environmental parameters, and welder skill parameters to output quality risk levels and welding parameter adjustment suggestions.

[0011] The quality result archiving module is used to integrate the risk records and adjustment instructions output by the pre-control module, as well as the UT / RT flaw detection results after welding is completed, and to establish a structured archive according to the hierarchy of "ship section - welding station - welder", which is then stored in the shipbuilding MES system; it provides data support for subsequent quality problem tracing (such as welding parameters and welder information associated with defects in a certain section), and also provides a data source for the historical data query function of the control terminal interaction module;

[0012] The control terminal interaction module is used to display welding parameter curves, risk levels (green / yellow / red pop-up prompts), and adjustment instructions in real time on the touch screen, allowing welders to intuitively obtain operation guidance and perform adjustments. The back-end associated quality result archiving module allows managers to log in with a password to query historical welding quality records and statistical reports for a specific segment or welder. It serves as both a real-time control operation window and a query entry point for post-event review, ensuring that control instructions are implemented and data is traceable.

[0013] Preferably, the specific working steps of the multi-source feature real-time alignment unit are as follows:

[0014] a1. Collect multi-dimensional raw data: Obtain welding electrical parameters and environmental parameters through the welding data acquisition module; obtain welder skill parameters through the welder identity authentication module;

[0015] b1. Timestamp calibration: Based on the sampling timestamp of the welding current, linear interpolation correction is performed on the molten pool temperature and ambient temperature and humidity data to eliminate sensor sampling delay and ensure that the data under the same timestamp correspond one-to-one, with a calibration error ≤ ±20ms;

[0016] c1. Dimensional unification: Map parameters of different dimensions to the [0,1] interval to eliminate the influence of dimensional differences on subsequent coupling analysis;

[0017] d1. Invalid data removal: Remove invalid data before and after the welding equipment is started and stopped, as well as abnormal values ​​caused by environmental sensor failures, and retain valid welding process data.

[0018] Preferably, the specific working steps of the welding quality risk level mapping unit are as follows:

[0019] a2. Construct a multi-factor weighted model: Based on pre-input statistical data of ship welding defects (such as "abnormal welding current" contributing 40% of porosity defects, "high ambient humidity" contributing 25%, "low welder level" contributing 15%, and "high wind speed" contributing 20%), assign weights to each parameter: welding current ω1 = 0.4, arc voltage ω2 = 0.3, ambient humidity ω3 = 0.25, welder historical pass rate ω4 = 0.3, and wind speed ω5 = 0.2;

[0020] b2. Calculate the real-time risk coefficient R: R = ω1 × I norm +ω2×U norm +ω3×H norm +(1-Q norm )×ω4+ω5×V norm , where I norm Normalized welding current; U norm Normalized arc voltage; H norm Normalized ambient humidity; Q norm Normalized historical pass rate for welders; V norm Normalized wind speed;

[0021] c2. Risk level classification: R < 0.3 is green low risk (no defect risk), 0.3 ≤ R < 0.6 is yellow medium risk (potential porosity / slag inclusion), and R ≥ 0.6 is red high risk (high probability of non-fusion / cracks);

[0022] d2. Risk source tracing: If R≥0.3, locate the 1-2 parameters with the largest weights (e.g., ω1×I). norm If the value is the highest, then "abnormal welding current" is determined to be the primary risk source.

[0023] Preferably, the specific working steps of the dynamic adjustment instruction generation unit are as follows:

[0024] a3. Establish a "risk source-adjustment parameter" mapping library: Based on preset ship welding process standards, store the adjustment rules corresponding to different risk sources;

[0025] b3. Refined instruction optimization: The adjustment range is optimized by combining the deviation of real-time welding parameters. For example, if the actual value of welding current is 420A and the process standard value is 350A (deviation 20%), then the adjustment instruction is "reduce the current to 350-385A (process value ±10%)".

[0026] c3. Command Priority Sorting: If there are multiple risk sources, the command is output according to the priority of "parameters affecting welding strength (such as current / voltage) > environmental parameters > personnel parameters". For example, if there are both "abnormal current" and "high humidity", the current adjustment command is output first.

[0027] d3. Instruction verification feedback: After adjustment, continuously monitor the risk coefficient R. If R drops to the target range within 5 seconds (medium risk → R < 0.3, high risk → R < 0.6), the instruction is deemed valid; if R does not drop, retrieve the mapping library again to optimize the instruction.

[0028] Preferably, a method for controlling the quality of ship welding, using the aforementioned ship welding quality control system, includes the following steps:

[0029] S1. Welder Identification: Welders are authenticated via RFID chips, and the system reads their skill parameters;

[0030] S2. Multi-dimensional data acquisition: Welding data acquisition module and environmental sensor synchronously acquire welding parameters and environmental parameters;

[0031] S3. Creative Pre-control Processing: The pre-control module completes multi-source feature alignment, risk level mapping, and dynamic instruction generation;

[0032] S4. Real-time interactive feedback: The control terminal displays the risk level and adjustment instructions, and the welder executes the adjustments;

[0033] S5. Quality Result Archiving: After welding is completed, the results of UT inspection are archived into the MES system.

[0034] Compared with existing technologies, the beneficial effects of this invention are as follows: In traditional control, welders often rely on experience to adjust parameters (such as "adjusting to a lower current if it is too high"), lacking quantitative basis, and are prone to new defects due to excessive adjustment (such as failure to fuse due to a sudden drop in current). The dynamic adjustment instruction generation unit of this invention is based on a "risk source-adjustment parameter" mapping library and real-time deviation, outputting instructions with specific parameter ranges (such as "humidity exceeds 80% RH, it is recommended to turn on the dehumidifier to 65%-70% RH"), and sorting them according to the priority of "intensity parameters > environmental parameters > personnel parameters", avoiding operational confusion under multiple risk sources.

[0035] This invention constructs a structured archive according to the hierarchy of "ship section - welding station - welder" through a quality result archiving module. It encrypts and stores risk records, adjustment instructions and final UT / RT flaw detection results in the MES system during the pre-control process. Managers can complete the historical data retrieval and statistical analysis of a certain section and a certain welder within 5 minutes through the management terminal interaction module.

[0036] This invention requires no hardware modification to existing ship welding equipment (such as CO2 gas shielded welding and submerged arc welding machines). It can be connected simply by integrating sensors (current and temperature sensors) with RFID readers, making it suitable for various welding positions on ships of different tonnages (from thousands of tons to hundreds of thousands of tons), such as decks, side plating, and bulkheads. At the same time, the system supports adjusting the "risk source-adjustment parameter" mapping library according to the process standards of shipbuilding enterprises (such as optimizing the current weight for high carbon steel welding) to meet personalized management and control needs.

[0037] In summary, this invention, through the core technological innovation of the pre-control module and the collaboration of multiple modules, not only solves the core pain points of traditional control, but also forms significant advantages in terms of quality stability, resource efficiency, and management convenience. It provides a practical and innovative solution for ship welding quality control, which is of great value for improving the overall quality of shipbuilding and promoting the intelligent transformation of the industry. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the overall system structure of the present invention;

[0039] Figure 2 This is a schematic diagram of the workflow of the multi-source feature real-time alignment unit of the present invention;

[0040] Figure 3 This is a schematic diagram of the workflow of the welding quality risk level mapping unit of the present invention;

[0041] Figure 4 This is a schematic diagram of the workflow of the dynamic adjustment instruction generation unit of the present invention;

[0042] Figure 5 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Please see Figure 1-5This invention provides a technical solution: a ship welding quality control system, including a welding data acquisition module, a welder identity authentication module, a pre-control module, a quality result archiving module, and a control terminal interaction module; the welding data acquisition module, through current, voltage, molten pool temperature sensors and workstation environment sensors integrated into the welding equipment, collects welding electrical parameters (80-500A current, 15-40V voltage, etc.), molten pool state (800-1500℃), and environmental parameters (temperature, humidity, wind speed) in real time, and outputs standardized raw data at a sampling frequency of ≥50Hz;

[0045] The welder identification module reads the welder's identity information (certificate level, preferred welding position, historical pass rate) through an RFID chip to verify whether the welder meets the skill requirements of the current workstation (e.g., vertical welding requires a Level III welder). If qualified, the welding equipment is unlocked. At the same time, the welder's skill parameters (especially the historical pass rate) are synchronized to the pre-control module.

[0046] The operation process for the welder identity authentication module is as follows:

[0047] Welders wear name tags with built-in RFID chips and approach the RFID reader of the welding equipment (reading distance ≤ 5cm);

[0048] The system reads the welder's ID, certificate level (e.g., Level II / Level III), preferred welding position (flat welding / vertical welding), and welding pass rate (e.g., 92%) from the chip.

[0049] If the welder's skill level meets the requirements of the current welding station (e.g., a Level III welder is required for vertical welding of the ship hull), the welding equipment will be unlocked; otherwise, the message "Insufficient permissions" will be displayed.

[0050] The pre-control module receives real-time process data from the welding data acquisition module and skill data from the welder identity authentication module. First, a multi-source feature real-time alignment unit performs data time calibration, dimension unification, and invalid data removal. Then, a welding quality risk level mapping unit, combined with a multi-factor weighting model, calculates risk coefficients, classifies risk levels (green / yellow / red), and locates risk sources. Finally, a dynamic adjustment instruction generation unit outputs precise adjustment suggestions based on the risk sources. The output risk level and adjustment instructions are synchronously pushed to the control terminal interaction module, while the process data is transmitted to the quality result archiving module. The pre-control module includes a multi-source feature real-time alignment unit, a welding quality risk level mapping unit, and a dynamic adjustment instruction generation unit. It can integrate welding electrical parameters, environmental parameters, and welder skill parameters to output quality risk levels and welding parameter adjustment suggestions.

[0051] The quality results archiving module integrates the risk records and adjustment instructions output by the pre-control module, as well as the UT / RT flaw detection results after welding is completed. It establishes structured archives according to the hierarchy of "ship section - welding station - welder" and stores them in the shipbuilding MES system. This provides data support for subsequent quality problem tracing (such as welding parameters and welder information associated with defects in a certain section) and also provides a data source for the historical data query function of the control terminal interaction module.

[0052] The control terminal interaction module is used to interact with the pre-control module, displaying welding parameter curves, risk levels (green / yellow / red pop-up prompts), and adjustment instructions in real time on the touch screen. This allows welders to intuitively obtain operation guidance and perform adjustments. The back-end associated quality result archiving module allows managers to log in with a password to query historical welding quality records and statistical reports for a specific segment or welder. It serves as both a real-time control operation window and a query entry point for post-event review, ensuring that control instructions are implemented and data is traceable.

[0053] Furthermore, the specific working steps of the multi-source feature real-time alignment unit are as follows:

[0054] a1. Collect multi-dimensional raw data: Obtain welding electrical parameters and environmental parameters through the welding data acquisition module; obtain welder skill parameters through the welder identity authentication module;

[0055] b1. Timestamp calibration: Based on the sampling timestamp of the welding current, linear interpolation correction is performed on the molten pool temperature and ambient temperature and humidity data to eliminate sensor sampling delay and ensure that the data under the same timestamp correspond one-to-one, with a calibration error ≤ ±20ms;

[0056] c1. Dimensional unification: Map parameters of different dimensions to the [0,1] interval to eliminate the influence of dimensional differences on subsequent coupling analysis;

[0057] d1. Invalid data removal: Remove invalid data before and after the welding equipment is started and stopped, as well as abnormal values ​​caused by environmental sensor failures, and retain valid welding process data.

[0058] Specifically, the multi-source feature real-time alignment unit is used to solve the problems of "time asynchrony and inconsistent units" in data from different sensors, providing standardized data for subsequent risk analysis. This will be explained with specific data examples:

[0059] a1. Data Acquisition: Simultaneously acquire welding current (e.g., 380A), arc voltage (28V), molten pool temperature (1200℃), ambient temperature (25℃), humidity (75% RH), wind speed (2.5m / s), and welder qualification rate (92%).

[0060] b1. Time calibration: The timestamp of the welding current is "1695000000.000s", and the timestamp of the molten pool temperature is "1695000000.015s". The temperature data is calibrated to "1695000000.000s" through linear interpolation to ensure synchronization;

[0061] c1. Dimensional uniformity:

[0062] Welding current 380A: process standard value 350A, normalized I nirm =(380-350) / (500-350)=0.2 (the proportion of the portion exceeding the standard to the adjustable range);

[0063] Ambient humidity 75% RH: Normalized H norm =75% / 90%≈0.83;

[0064] Welder qualification rate 92%: Normalized Q norm =92% / 100% = 0.92;

[0065] d1. Invalid data removal: Remove data from the preheating stage of the welding equipment (current < 80A) and retain valid data from the normal welding stage (current 80-500A).

[0066] Furthermore, the specific working steps of the welding quality risk level mapping unit are as follows:

[0067] a2. Construct a multi-factor weighted model: Based on pre-input statistical data of ship welding defects (such as "abnormal welding current" contributing 40% of porosity defects, "high ambient humidity" contributing 25%, "low welder level" contributing 15%, and "high wind speed" contributing 20%), assign weights to each parameter: welding current ω1 = 0.4, arc voltage ω2 = 0.3, ambient humidity ω3 = 0.25, welder historical pass rate ω4 = 0.3, and wind speed ω5 = 0.2;

[0068] b2. Calculate the real-time risk coefficient R: R = ω1 × I norm +ω2×U norm +ω3×H norm +(1-Q norm )×ω4+ω5×V norm , where I norm Normalized welding current; U norm Normalized arc voltage; H norm Normalized ambient humidity; Q norm Normalized historical pass rate for welders; V norm Normalized wind speed;

[0069] c2. Risk level classification: R < 0.3 is green low risk (no defect risk), 0.3 ≤ R < 0.6 is yellow medium risk (potential porosity / slag inclusion), and R ≥ 0.6 is red high risk (high probability of non-fusion / cracks);

[0070] d2. Risk source tracing: If R≥0.3, locate the 1-2 parameters with the largest weights (e.g., ω1×I). norm If the value is the highest, then "abnormal welding current" is determined to be the primary risk source.

[0071] The following explanation uses specific data:

[0072] a2. Weighting: Based on the statistical data of ship welding defects, assign weights ω1 = 0.4 (current), ω2 = 0.3 (voltage), ω3 = 0.25 (humidity), ω4 = 0.3 (welder qualification rate), and ω5 = 0.2 (wind speed);

[0073] b2. Calculate the risk factor:

[0074] Assuming an arc voltage of 28V (process standard 26V), normalized U norm = (28-26) / (40-26)≈0.14;

[0075] Wind speed 2.5 m / s, normalized V norm =2.5 / 10 = 0.25;

[0076] R=0.4×0.2+0.3×0.14+0.25×0.83+(1-0.92)×0.3+0.2×0.25≈0.08+0.042+0.2075+0.024+0.05=0.4035;

[0077] c2. Risk level determination: 0.3≤R<0.6→Yellow, medium risk, potential defect is porosity;

[0078] d2. Risk source tracing: ω3×H norm =0.25×0.83=0.2075 (highest weighting), thus "high ambient humidity" is determined to be the main source of risk.

[0079] Furthermore, the specific working steps of the dynamically adjusted instruction generation unit are as follows:

[0080] a3. Establish a "risk source-adjustment parameter" mapping library: Based on preset ship welding process standards, store the adjustment rules corresponding to different risk sources;

[0081] b3. Refined instruction optimization: The adjustment range is optimized by combining the deviation of real-time welding parameters. For example, if the actual value of welding current is 420A and the process standard value is 350A (deviation 20%), then the adjustment instruction is "reduce the current to 350-385A (process value ±10%)".

[0082] c3. Command Priority Sorting: If there are multiple risk sources, the command is output according to the priority of "parameters affecting welding strength (such as current / voltage) > environmental parameters > personnel parameters". For example, if there are both "abnormal current" and "high humidity", the current adjustment command is output first.

[0083] d3. Instruction verification feedback: After adjustment, continuously monitor the risk coefficient R. If R drops to the target range within 5 seconds (medium risk → R < 0.3, high risk → R < 0.6), the instruction is deemed valid; if R does not drop, retrieve the mapping library again to optimize the instruction.

[0084] The following explanation is based on the data mentioned above:

[0085] a3. Retrieve the mapping library: The basic instruction corresponding to the risk source "high ambient humidity (H>80% RH)" is "turn on the dehumidifier";

[0086] b3. Optimization adjustment range: The current humidity is 75%RH (close to the threshold of 80%RH). No need to stop the machine. The optimized instruction is "Turn on the dehumidifier at the workstation and control the humidity at 65%-70%RH".

[0087] c3. Priority sorting: Currently, there is only one major risk source, "high humidity", so this instruction is output directly;

[0088] d3. Verification Feedback: 5 seconds after the dehumidifier was turned on, the humidity dropped to 68% RH. norm =68% / 90%≈0.76, R was recalculated to ≈0.36 (still medium risk). Further inspection revealed that the welding current was still too high (380A). The supplementary instruction was "reduce the welding current to 350-360A". After adjustment, R dropped to 0.28 (low risk). The instruction was verified to be effective.

[0089] Furthermore, a method for controlling the quality of ship welding, using a ship welding quality control system, includes the following steps:

[0090] S1. Welder Identification: Welders are authenticated via RFID chips, and the system reads their skill parameters;

[0091] S2. Multi-dimensional data acquisition: Welding data acquisition module and environmental sensor synchronously acquire welding parameters and environmental parameters;

[0092] S3. Creative Pre-control Processing: The pre-control module completes multi-source feature alignment, risk level mapping, and dynamic instruction generation;

[0093] S4. Real-time interactive feedback: The control terminal displays the risk level and adjustment instructions, and the welder executes the adjustments;

[0094] S5. Quality Result Archiving: After welding is completed, the results of UT inspection are archived into the MES system.

[0095] This invention addresses the core pain points of traditional ship welding quality control, offering significant advantages in quality assurance, resource efficiency, operational standardization, management convenience, and implementation costs, as detailed below:

[0096] At the quality pre-control level, traditional management methods suffer from fragmented data on welding electrical parameters, ambient temperature and humidity, and welder skills, making it impossible to identify risks arising from the coupling of multiple factors. Furthermore, problems can only be detected after the fact through post-construction flaw detection. This invention relies on a core pre-control module. First, it achieves data time synchronization and dimensional unification through a multi-source feature real-time alignment unit. Then, it constructs a multi-factor weight model through a welding quality risk level mapping unit to accurately calculate risk coefficients and predict defects. It is particularly effective in preventing and controlling critical defects such as incomplete fusion and porosity in complex work positions such as vertical and horizontal welding of ship hulls, directly ensuring the structural strength of the hull and the safety of ship navigation.

[0097] In terms of resource and schedule optimization, rework due to defects in the traditional model not only wastes materials such as steel and welding wire, but also leads to schedule delays. This invention, through a dynamically adjusted instruction generation unit, immediately outputs intervention instructions with specific parameter ranges (such as "current exceeds standard by 20%, it is recommended to reduce to 350-385A") when a risk occurs, and verifies the effectiveness of the instruction within 5 seconds, achieving "risk-based intervention" and effectively avoiding rework afterward.

[0098] At the operational and management level, traditional welders rely on experience to adjust parameters, achieving an accuracy rate of only around 60%, which easily leads to new defects. Furthermore, quality data is stored in a fragmented manner, requiring cross-platform integration for traceability, and retrieving single-segment issues can take 2-3 hours. The instruction output mechanism of this invention increases welder operational accuracy to over 90%. Simultaneously, the quality result archiving module constructs structured files according to "ship section - welding station - welder," allowing managers to complete data retrieval within 5 minutes via a control terminal. This reduces human error, clarifies quality responsibility, and improves management efficiency.

[0099] In terms of practical application, this invention does not require modification of existing CO2 welding, submerged arc welding and other welding equipment. It only requires the integration of sensors and RFID readers for docking. Furthermore, deployment at a single workstation takes only 1-2 days. It is compatible with various welding workstations on ships of different tonnages, significantly reducing the threshold for intelligent transformation of shipbuilding enterprises, facilitating rapid promotion and application, and helping to upgrade the industry's welding quality control.

[0100] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A ship welding quality control system, characterized in that, It includes a welding data acquisition module, a welder identity authentication module, a pre-control module, a quality result archiving module, and a management terminal interaction module; the welding data acquisition module collects welding electrical parameters, molten pool status, and environmental parameters in real time through current, voltage, molten pool temperature sensors and station environment sensors integrated into the welding equipment, and outputs standardized raw data at a sampling frequency of ≥50Hz; The welder identification module reads the welder's identity information through an RFID chip to verify whether the welder meets the skill requirements of the current workstation. If the welder meets the requirements, the welding equipment is unlocked. At the same time, the welder's skill parameters are synchronized to the pre-control module. The pre-control module receives real-time process data from the welding data acquisition module and skill data from the welder identity authentication module. First, the multi-source feature real-time alignment unit completes data time calibration, dimension unification, and invalid data removal. Then, the welding quality risk level mapping unit calculates the risk coefficient, classifies the risk level, and locates the risk source by combining the multi-factor weight model. Finally, the dynamic adjustment instruction generation unit outputs precise adjustment suggestions based on the risk source. The risk level and adjustment instructions output by the module are simultaneously pushed to the control terminal interaction module, while the process data is transmitted to the quality result archiving module; the pre-control module includes a multi-source feature real-time alignment unit, a welding quality risk level mapping unit, and a dynamic adjustment instruction generation unit; The quality result archiving module is used to integrate the risk records and adjustment instructions output by the pre-control module, as well as the UT / RT flaw detection results after welding, and to establish a structured archive according to the hierarchy of "ship section - welding station - welder", which is then stored in the shipbuilding MES system; this provides data support for subsequent quality problem tracing and also provides a data source for the historical data query function of the control terminal interaction module; The control terminal interaction module is used to display welding parameter curves, risk levels, and adjustment instructions in real time on the touch screen, allowing welders to intuitively obtain operation guidance and perform adjustments. The back-end associated quality result archiving module allows managers to log in with a password to query historical welding quality records and statistical reports for a specific segment or welder. It serves as both a real-time control operation window and a query entry point for post-event review, ensuring that control instructions are implemented and data is traceable.

2. The ship welding quality control system according to claim 1, characterized in that: The specific working steps of the multi-source feature real-time alignment unit are as follows: a1. Collect multi-dimensional raw data: Obtain welding electrical parameters and environmental parameters through the welding data acquisition module; obtain welder skill parameters through the welder identity authentication module; b1. Timestamp calibration: Based on the sampling timestamp of the welding current, linear interpolation correction is performed on the molten pool temperature and ambient temperature and humidity data to eliminate sensor sampling delay and ensure that the data under the same timestamp correspond one-to-one, with a calibration error ≤ ±20ms; c1. Dimensional unification: Map parameters of different dimensions to the [0,1] interval to eliminate the influence of dimensional differences on subsequent coupling analysis; d1. Invalid data removal: Remove invalid data before and after the welding equipment is started and stopped, as well as abnormal values ​​caused by environmental sensor failures, and retain valid welding process data.

3. The ship welding quality control system according to claim 1, characterized in that: The specific working steps of the welding quality risk level mapping unit are as follows: a2. Construct a multi-factor weighted model: Based on pre-input statistical data of ship welding defects, assign weights to each parameter: welding current ω1 = 0.4, arc voltage ω2 = 0.3, ambient humidity ω3 = 0.25, welder historical pass rate ω4 = 0.3, and wind speed ω5 = 0.2; b2. Calculate the real-time risk coefficient R: R = ω1 × I norm +ω2×U norm +ω3×H norm +(1-Q norm )×ω4+ω5×v norm , where I norm Normalized welding current; U norm Normalized arc voltage; H norm Normalized ambient humidity; Q norm Normalized historical pass rate for welders; V norm To normalize wind speed; c2. Risk level classification: R < 0.3 is green (low risk), 0.3 ≤ R < 0.6 is yellow (medium risk), and R ≥ 0.6 is red (high risk). d2. Risk source tracing: If R≥0.3, locate the 1-2 parameters with the largest weight.

4. The ship welding quality control system according to claim 1, characterized in that: The specific working steps of the dynamic adjustment instruction generation unit are as follows: a3. Establish a "risk source-adjustment parameter" mapping library: Based on preset ship welding process standards, store the adjustment rules corresponding to different risk sources; b3. Command Refinement Optimization: Optimize the adjustment range by combining real-time welding parameter deviations; c3. Command Priority Sorting: If multiple risk sources exist, commands are output according to the priority order of "parameters affecting welding strength > environmental parameters > personnel parameters"; d3. Instruction verification feedback: After adjustment, continuously monitor the risk coefficient R. If R drops to the target range within 5 seconds, the instruction is deemed valid; if R does not drop, re-fetch the mapping library to optimize the instruction.

5. A method for quality control of ship welding, characterized in that, The application of the ship welding quality control system according to any one of claims 1-4 includes the following steps: S1. Welder Identification: Welders are authenticated via RFID chips, and the system reads their skill parameters; S2. Multi-dimensional data acquisition: Welding data acquisition module and environmental sensor synchronously acquire welding parameters and environmental parameters; S3. Creative Pre-control Processing: The pre-control module completes multi-source feature alignment, risk level mapping, and dynamic instruction generation; S4. Real-time interactive feedback: The control terminal displays the risk level and adjustment instructions, and the welder executes the adjustments; S5. Quality Result Archiving: After welding is completed, the results of UT inspection are archived into the MES system.

Images